Statistical Test for Anomaly Detections by Variational Auto-Encoders
This addresses the need for reliable anomaly detection in high-stakes domains like medical diagnosis, offering a statistically grounded method, though it is incremental as it builds on existing VAE-based approaches.
The study tackled the problem of ensuring reliability in anomaly detection using Variational Autoencoders by proposing the VAE-AD Test, which quantifies statistical reliability with p-values to control false detection probabilities, and demonstrated its validity through numerical experiments and brain image applications.
In this study, we consider the reliability assessment of anomaly detection (AD) using Variational Autoencoder (VAE). Over the last decade, VAE-based AD has been actively studied in various perspective, from method development to applied research. However, when the results of ADs are used in high-stakes decision-making, such as in medical diagnosis, it is necessary to ensure the reliability of the detected anomalies. In this study, we propose the VAE-AD Test as a method for quantifying the statistical reliability of VAE-based AD within the framework of statistical testing. Using the VAE-AD Test, the reliability of the anomaly regions detected by a VAE can be quantified in the form of p-values. This means that if an anomaly is declared when the p-value is below a certain threshold, it is possible to control the probability of false detection to a desired level. Since the VAE-AD Test is constructed based on a new statistical inference framework called selective inference, its validity is theoretically guaranteed in finite samples. To demonstrate the validity and effectiveness of the proposed VAE-AD Test, numerical experiments on artificial data and applications to brain image analysis are conducted.